With regard to bumblebee 'migration', it isn't about individuals physically relocating. Rather, what is expected is that as temperatures warm, species sensitive to temperature will start to have better reproductive success on one end of their range that favors them, and reduced success at the other end of the range. Over time the favorable edge expands and unfavorable edge shrinks and the whole population moves over many generations. But this presumes that the necessary habitat exists in the direction the population is moving. If you reach a northern forest with no wildflowers for hundreds of miles, now what?
There's really a dearth of good long term data on wild bees:
Jennifer Sass, of the Natural Resource Defense Council, acknowledged in a webinar put on by Northeastern IPM Center that there is no data available showing the decline of wild bees, saying environmental groups can only “presume” that wild bees are in decline. But the science doesn’t back up that presumption.
The advantage to this new narrative is that there simply are no good statistics on wild bees. We don’t even know how many species of wild bees there are, let alone their numbers or the population trends over time. Except for several bumblebee species that have collapsed due to disease, no baseline data exist on indigenous bee populations in this country and how they are changing over time–just the kind of black box science that invites advocacy scare mongering.
There is every reason to believe, however, that the new crisis is just as thinly supported by science as the old one. A 2013 study published in the Proceedings of the National Academy of Sciences analyzed U.S. native bee populations over a 140-year period. Only three of 187 native species had declined steeply, almost certainly due to a pathogen and not any pesticide. In Europe, wild bee populations have been in decline for nearly seventy years, but in recent decades the decline has slowed and in some cases turned around, even as the use of neonics has skyrocketed
A recent three-year study published in Nature, conducted by 58 scientists around the world, found that the wild bee species that pollinate crops (and which would therefore come into the greatest contact with neonics) are flourishing.
Sam Droege, a wild bee specialist at the U.S. Geological Survey who has been tasked with performing the first comprehensive survey ever of wild bee populations in the U.S., has said his surveys so far indicate that most wild bees are doing just fine.
Via Beemageddon? As hysteria over honey bees recedes, anti-neonic narrative refocuses on wild bees
Long term data is important because bee populations can naturally double or halve in just one year:
Year-to-year shifts in bee abundance for three tropical and five temperate bee censuses were comparable. In short studies (2-4 years) and during longer studies (17-21 years), 59 species that included solitary, social, and highly social bees had mean abundances that varied by factors of 2.06 for temperate bees and 2.16 for tropical bees. “Normal” bee populations commonly halved or doubled in 1-yr intervals. Longer term data are only available for the tropics. Stochastic variation and limitations of monitoring methods suggest that minimum series of four years (i.e., three intervals) of several counts during the active season may demonstrate genuine trends. Longer term, continuous studies are still needed for meaningful insights on pollinator population shifts in nature.
Via Ups and Downs in Pollinator Populations: When is there a Decline?
The author explains a little more the causes of doubling/halving - more rain means more flowers which make more nectar that makes more bees:
One benefit of doing long-term studies is that I see what happens when an El Niño year comes in the tropics, which causes sustained and super-productive flowering and feeds a lot more bees than the normal,” he said. “This makes populations go up and then go down—they're supposed to do that. After a year or two of big decline people will start saying Henny Penny the sky is falling, but you can't predict anything on the basis of one or two years of study. Stability is not the norm, not here or anywhere else.
Via Smithsonian's Bee Man Delivers Up Some Advice for Dealing with Colony Collapse Disorder